Should Surveillance Cameras and Facial Recognition Software Be Used to Monitor Public Spaces?

Combine the  now-ubiquitous state-of-the-art high-resolution CCTV surveillance cameras with facial recognition software, which analyzes a digital image of a person and tries to match his or her features with a database of subjects, and you’ve got a system that potentially could enable law enforcement and intelligence agencies to spot criminals or terrorists in public places before they have a chance to act. Or maybe you’ve got a tool for authoritarian social control that’s so creepy it belongs in a Philip K. Dick novel. Or maybe both.

The casino industry has been using surveillance cameras and facial recognition software to spot known card-counters and other undesirables for years, and officials in the London suburb of Newham noted that crime declined by 15 percent after they installed such a system in the late 1990s, according to an article from the Independent, a UK newspaper. German and Australian immigration officials now employ the technology to screen travelers arriving in their countries. In the U.S., early experiments with facial recognition surveillance of crowds yielded uneven results. At the 2001 Super Bowl in Tampa, for example, facial recognition software scanned everyone passing through the turnstiles and came up with 19 possible matches with mug shots of known criminals — but according to a ZDNet.com article, many of those turned out to be false hits.

Since then, however, advances such as three-dimensional face recognition and digital image skin correlation, which analyzes an individual’s unique motion of skin pores when making a facial expression, have emerged. In tests, the DISC method has enabled researchers to identify even subjects who were wearing heavy makeup, and some believe it ultimately may rival the accuracy of that 1800s-vintage biometric gold standard, fingerprint identification.  Given the U.S. government’s post- Sept. 11 tendency to cast a wide net in  surveillance — such as the National Security Agency’s reported amassing of a database of 1.9 trillion phone calls made in the U.S. — it’s not surprising that development of facial identification technology is part of its Homeland Security strategy.  As InformationWeek reported in 2005, In-Q-Tel, a CIA-backed venture capital group, has invested in A4Vision Inc., a maker of 3-D facial-scanning and -recognition software and equipment.

Next-generation facial recognition surveillance may not only discern a subject’s identity, but possibly his or her emotional state as well. As this article in Technology Review details, the Japanese company Omron recently demonstrated the abilities of its facial recognition software to analyze a subject’s smile, based on how characteristics such as mouth wideness, eye narrowing and facial wrinkles compare to a database of 10,000 different grins.

It’s not hard to imagine how powerful of a tool facial recognition surveillance could be for law enforcement and intelligence agencies. By imbedding facial recognition software within individual nodes of a video surveillance network, it might be possible to spot and identify a known criminal or terrorism suspect and then automatically follow him with the camera lens — and then continue the monitoring with cameras located elsewhere in the city as he made his way down its streets. By factoring in proximity to likely targets, such a system might give police the ability to swoop in and stop a crime the second it occurs — or to intercept a suicide bomber before he can detonate his weapon.

Unfortunately, it’s also not hard to imagine how a malevolent government could use such technology to stifle protest marches before they can occur, or to track the movements of political dissidents, or to gather embarrassing information about legislators’ extramarital activities that can be used to blackmail them into meek compliance with presidential policies.

So what do you think? If facial recognition technology is perfected, should it be used to monitor public spaces? Join the debate here.

Comments

[Becky T.]

I think the cameras are good because they are add safety
features to life. I mean,-for crying out loud, we don't
have to be afraid of them unless of coars we are trying to
commit a crime, which is something most people enjoy doing.

[Mothra]

It just seems like would be no place that you could be anonymous or go about your business privately anymore. I don't like that.

[Daniel Klaisner]

Ok heres the deal. With this technology finding a criminal or terrorist would be a lot easier and can be done in minutes. Or lets say your child gets kidnapped, This technology would find him or her almost imidiatly. And as far as privacy goes, nobody will be looking at you at all times. There are 6 billion people on this planet they cant and wont watch everybody. This is a security system not a spy system. I understand how its uncomfortable but the benifits are incredable. You never know when you will need to find your missing child

[Mothra]

I can understand how this technology might be used in a positive way. However, I also can imagine all the negative ways in which it could be used. If the government wanted to interfere with an opposition party, a labor union or an environmental activist group, it would be possible to track known members' movements and to identify and track new members as well. The ability to track people and intercept them before they get to a destination would be great for disrupting protest rallies or sit-ins before they can even happen. Given the extent to which our government already is abusing its surveillance powers, do we want to give officials an even greater power?

[Daniel Klaisner]

Good point, I wonder if there was a way to get it out of the governments control. If not we just have to weigh the good with the bad and see whats better for society as a whole.

[Patrick]

fyi...I just exchanged e-mails with a spokesperson for Eptascape (http://www.eptascape.com), a company whose surveillance software not only tracks subjects but actually analyzes their behavior, with the intent of spotting suspicious activities. Hopefully, an expert from the company will join in the discussion to tell us more about what they can do.

[Patrick]

I think Daniel raises some excellent points about the potential value of this technology, if it's used for positive purposes. I'm wondering if the civil liberties concerns might be addressed by setting legal limits on how such surveillance would be conducted and how the information could be used. Anyone out there with an opinion on whether such restrictions would be workable? What should the legal limits be? Also, should surveillance networks with video analytics be limited to government use, or should the private sector be allowed to do it also? Should the government regulate private use of this technology?

[Evan Hendricks--publisher, Privacy Times]

Surveillance camera nets and next-generation facial recognition software pose serious threats to privacy. First the accuracy and integrity of these systems have never been audited or tested. If relied upon, they will result in the same type of false positives and inaccuracies we've seen in other realms, such as identity theft and credit reporting. 2nd, this information ultimately will be 'databased', meaning there will be a secret set of files of people whose faces are caught, including those misidentified by the faulty software. Secret files are an anethema to privacy, as the Stasi proved in East Germany. People wont know about them, and certainly wont have access and correction rights.

Third, with all new invasive technologies, there is usually companies/contractors that stand to profit -- and often are found to be the force behind the push for their adoption. More sunshine is needed for these companies, as some of them are modern snake-oil salesmen.

[Patrick]

Oops...when I posted Evan's comments for him, I misidentified him as the publisher--he's actually the editor of Privacy Times. Check out the website http://www.privacytimes.com/
for more discussion of privacy, civil liberties and technology.